Emerging Trends in Computing |
Foundation of Computer Science USA |
ETC2016 - Number 3 |
March 2017 |
Authors: Vishakha B. Pawar, J. R. Mankar |
94bd5c1b-1bde-42d4-8532-9a42b09c8432 |
Vishakha B. Pawar, J. R. Mankar . Matching and Retrieval of Near Duplicate Images. Emerging Trends in Computing. ETC2016, 3 (March 2017), 10-13.
Using some transformation on original images, images can be modified which forms near duplicate images. Using the size and length of image, image represented and the length is variable with number of patches in the image. Spatially adjacent and similar pixels are used to form clusters. Patch means a form of clusters. Image representation and image similarity measurement are two major issues in image matching. The proposed method extracts patches from given image and represents by variable length signature. In the detection of duplicate natural images the signatures are use. Then take a decision for images are duplicates or not. The retrieval of near duplicate images means a query image and datasets are given then the near duplicate images are retrieving. Similarity is computed between two images, which can be done on variable length signature to improve the effectiveness.